Tuesday, September 29, 2009

New IR Poll: 7-pt Conservative Lead

Ipsos-Reid has released a new poll, taken between September 22 and September 24 and involving 1,001 Canadians.

It further confirms the findings of the other polling houses. In fact, the polls that have been released since last Thursday have been remarkably consistent. Here are the national results:

Conservatives - 37%
Liberals - 30%
New Democrats - 14%
Bloc Quebecois - 9%
Greens - 9%

This really is what we've been seeing all week. The regional results also echo most of what the other polls have found.

In British Columbia, the Conservatives lead with 36% but the Liberals are doing very well at 32%. The NDP, at 22%, is falling behind. The Greens also struggled in this province with 9%.

Alberta has 61% for the Conservatives, 16% for the Liberals, 13% for the Greens, and 10% for the NDP.

The Prairies has a very good 53% for the Conservatives, followed by 23% for the Liberals and 15% for the NDP.

The Conservatives have again polled over 40% in Ontario, with 43%. The Liberals are 10-points behind at 33%. The NDP is at 14%. This has been a trend we've seen all week. The Conservative lead is anywhere from five to ten points over the Liberals in the most important province in the country.

In Quebec, the Bloc is back to where it usually is at 37%. The Liberals are at 28% and the Conservatives at 17%. The NDP is at 10%. These, too, are surprisingly consistent results.

In Atlantic Canada, the Liberals are far ahead with 47%. This is something we've seen as well. Clearly, while the Liberals seem to be falling in Ontario and a little in Quebec, the political manoeuvres in Ottawa have pushed them well ahead in Atlantic Canada. This is confirmed by the leadership and "election yes-or-no" polls we've seen in the region. The Conservatives are at 33%, a good result, while the NDP is at 13%, disastrous.

This poll would result in the following seat totals:

Conservatives - 146
Liberals - 95
Bloc Quebecois - 49
New Democrats - 18

Again, the Tories still can't get to majority territory. They posted excellent numbers in Alberta, the Prairies, Ontario, and Atlantic Canada. But as I've said on numerous occasions, you can't win a majority without Quebec.

This electoral result would also be very bad on Jack Layton's political career. One wonders if Michael Ignatieff would be safe as leader if he increased his caucus by 20 MPs.

The Liberals have put forward a short and sweet non-confidence motion, to be voted on at the end of the week. Gilles Duceppe has already committed his party to it, but Layton has decided to support the government. So, that pretty much bars an election this fall. The NDP will have to stick to this new principle of waiting for the EI-reform to become law, meaning the government is safe for at least a month. Then when November rolls around, the idea of sending people to the polls during the Christmas season will likely keep us out of an election as well. Then in January, an election would interfere with the Vancouver Olympics, which gives the government life until the end of February.

That means it is extremely unlikely that we would have an election until March. And, at that point, I'm sure someone will think of an excuse not to send people to the polls. "It's Spring! We can't ask people to vote when the flowers are blooming."

18 comments:

  1. An outlier no more.

    ReplyDelete
  2. Why don't the liberals accept the fact they are the opposition party and will still be in opposition if we have an election.
    They have never accepted defeat in the last three elections. Start rebuilding and start coming up with programs or ideas instead of We were robbed of access to the trough.
    They are lost without the ability to get rid of MPs with appointments, patronage and brown paper bags.
    Quit playing games and grow up.

    ReplyDelete
  3. With every poll showing AB going CPC you still give it to Lib vs NDP?

    Do you have the special riding poll data to suggest the Liberals win that from the NDP or the CPC don't win it back?

    http://punditsguide.ca/riding_e.php?riding=1205

    Liberals stand to gain 16,000 votes to beat NDP in Edmonton Stratcona?

    ReplyDelete
  4. Who said anything about Edmonton-Strathcona?

    Every poll has shown the Liberals have grown their strength in Alberta by 50% or even 100%.

    ReplyDelete
  5. I'd like to remind everyone that as the blog operator I am responsible for the comments posted on this blog. So please don't post libelous comments like the one I just deleted.

    ReplyDelete
  6. One should google the results of the 2008 election results for Alberta.
    Duncan won by 442 votes, all incumbents won by several thousand votes over the liberals, most getting more than 50% of the votes cast. The average liberal votes/riding was around 5000, with the winner getting over 20,000 votes. Increasing their vote by 100% will still leave then around 10,000 votes behind.

    ReplyDelete
  7. Eric,

    I think we all just assumed Edmonton-Strathcona because that's the only Alberta seat the Conservatives don't have.

    Does your model tell you which riding the Liberals are supposed to win ? One of the last to fall was Landslide Annie's riding which is now more or less Edmonton Centre but its hard to see them winning that back.

    ReplyDelete
  8. I guess the overall math here may explain why all the exaggerating hysteria over 'the so respected' Coderre being dumped for having tried to stop the Toronto tinhorns from imposing "free grassroots democracy" in Quebec ridings.
    And notice how the media's jumped head over heels into it, conveniently giving Harper and his bogus claims a free pass...

    ReplyDelete
  9. Rockfish,

    my observation is months ago LPC were ahead in Ontario by 10% in Ontario.

    Today the gap has shrunk or show the CPC in the lead by 5%.

    The Polls do not have specific riding information but Ontario had many close ridings that have not changed hands but 1 AB still remain a Liberal seat.

    The Liberals are third place in AB behind the NDP and the gap is over 5,000+.

    Eric claims the increase of 50%-100% justifies his forecast.

    I pointed out the same problem with the North.

    He has somewhat corrected it.

    No other projection site has Liberals winning in AB or loss for the NDP as significant at the benefit of the Liberals.

    ReplyDelete
  10. MaryT,

    --- "The average liberal votes/riding was around 5000, with the winner getting over 20,000 votes. Increasing their vote by 100% will still leave then around 10,000 votes behind."

    Assuming the votes come from the Conservatives in an individual riding, increasing their vote by 100% would leave them 5,000 behind, not 10,000. Individual riding swings, with good candidates or unexpected events, are not unusual.

    Jesse,

    --- "Does your model tell you which riding the Liberals are supposed to win ?"

    No.

    CanadianSense,

    --- "The Liberals are third place in AB behind the NDP and the gap is over 5,000+."

    What does this mean?

    In the last election the NDP had 12.7%, the LPC had 11.4%, and the CPC had 64.6%.

    My poll aggregator now has the NDP at 11.4% (a reduction of more than 10%), the LPC at 17.5% (an increase of 54%), and the CPC at 61.1% (a reduction of more than 5%).

    If we applied those changes directly to last year's vote (not a good way to do things, but all we have for this discussion), we get in Edmonton-Strathcona:

    CPC - 40%
    NDP - 38%
    LPC - 14%

    Edmonton-Centre:

    CPC - 47%
    LPC - 42%
    NDP - 14%

    A five-point gap using this method is absolutely winnable for the Liberals.

    --- "I pointed out the same problem with the North. He has somewhat corrected it."

    Your question had nothing to do with it. The Conservatives have won back a seat in the North because their polling has improved.

    ReplyDelete
  11. Eric,

    you don't have any local polling information.

    Your projection regarding the North?


    The 99% Polling is National, north is excluded. In those National Polls smaller subsets into provinces, region with +10% MOE for smaller regions leaving only ONT. QC most consistent with the largest sample size.
    You don't have riding by riding breakdown in 905,416,705 etc.

    You allow NDP to hold QC, lose AB, and Liberals hold 905, pickup up BC seats.

    In my opinion your seat projection is inconsistent without specific polling data into ridings as I can not find a second projection site that gives the Liberals 100+ seats (as you do consistenly)since June-September as the Polls have moved negatively for the Liberals.

    ReplyDelete
  12. --- "you don't have any local polling information."

    I don't need any. If we had local riding polls, we wouldn't need seat projections.

    --- "The 99% Polling is National, north is excluded. In those National Polls smaller subsets into provinces, region with +10% MOE for smaller regions leaving only ONT. QC most consistent with the largest sample size.
    You don't have riding by riding breakdown in 905,416,705 etc."

    This is non-sensical.

    -- "You allow NDP to hold QC, lose AB, and Liberals hold 905, pickup up BC seats."

    That's what the polls show.

    -- "I can not find a second projection site that gives the Liberals 100+ seats (as you do consistenly)since June-September as the Polls have moved negatively for the Liberals."

    The projection will drop below 100 Liberal seats if this trend continues. This projection is meant to express only consistent trends, not what people have told pollsters over the last few weeks. It is meant to show the results of an electoral campaign based on current trends.

    ReplyDelete
  13. Come on guys, Eric is doing a fine job of presenting polling data. Of course his model will have errors as it does NOT take into account specific peculiarities of certain ridings, where some ridings will relect an MP based on him and not the National Trend.

    Sure perhaps Eric should remove the 1 Liberal seat in AB, but what does it matter, since the CPC takes the 27 others.

    This site is a lot of fun, and Eric presents some of the most resonable trend analysis anywhere. So it is not necessary to nitpick on specific ridings. If you want to nitpick go to the Election Prediction Project http://electionprediction.org/

    ReplyDelete
  14. Thanks! Was starting to think I had only critics.

    ReplyDelete
  15. Eric,

    In my opinion you have put in a great deal of effort and I applaud those efforts.

    My personal opinion has nothing to do with my political preference but the difference in your projection number vs others in the last few months.

    I don't understand how QC NDP hold its seat, 905 stay Lib, Libs gain in BC while other projection sites have shown the movement as the polls change.

    The Liberals 37-29% since June, CPC 30-39% have moved up, NDP 12-17% in National Polls.

    Other projection sites have reflected the negative trend in liberal support with Polls in seats over -3% when they were ahead of the CPC in early June.

    Again I APPLAUD your incredible effort on your site.

    (I just don't understand how you do your seat projection does not match other sites more closely if your are all using the same pollling data)

    ReplyDelete
  16. We all have different methods. The biggest difference is that I don't use only recent polls. They swing more sharply according to what polls have been released in the past few weeks, while I am more conservative in taking new results at face-value.

    I'll point out that in the Quebec provincial election of 2008, my site was the closest.

    ReplyDelete
  17. Eric

    The National Polls have a MOE of +1%

    The CPC are within their 2008 numbers as are the NDP?

    The Liberals are polling 3% better than the October 2008, and you give them +30 seats making their votes highly "STRATEGIC" without being able to back up specific ridings since NO public company offers that polling insight.

    That is my observation vs other pollsters/projections sites who reflect the polls more closely without any "strategic" allocation into specific ridings for the benefit in (((((MY)))) opinion on behalf of Liberal votes.

    In conclusion I think you should be applauded for all your WORK and effort in this wonderful site.

    I just don't get the strategic voting pattern you infer in your projection.

    ReplyDelete
  18. -- "The National Polls have a MOE of +1%"

    All projectors can only work with the information they are provided with by the pollsters. When you take all of the polling results together, however, the MOE is minimal.

    -- "The CPC are within their 2008 numbers as are the NDP?"

    The CPC are within their 2008 numbers in only a few recent polls. The NDP are certainly not at their 18% total from last year, and are doing very bad in some important regions.

    But polls are polls, and my projection is based on voting behavior.

    -- "The Liberals are polling 3% better than the October 2008"

    Which represents a more than 11% increase, if you mean they are at 29%. My aggregator still has them at a little over 31%, representing a 20% increase.

    -- "and you give them +30 seats"

    The national vote is meaningless. It is their very good polling results in British Columbia, Quebec, and to a lesser extent in the Prairies, Alberta, Ontario and Atlantic Canada, that give the 30+ seats, coupled with the NDP doing worse.

    They are at 2006 levels, when they got 100+ seats.

    -- "without being able to back up specific ridings since NO public company offers that polling insight."

    No projector has that benefit, THAT'S WHY WE MAKE SEAT PROJECTIONS.

    -- "That is my observation vs other pollsters/projections sites who reflect the polls more closely"

    If they react too quickly to recent polls, their projection will suffer.

    -- "without any "strategic" allocation into specific ridings for the benefit in (((((MY)))) opinion on behalf of Liberal votes."

    I'm only using the polling data made available to me and comparing that to historical results. That's it, that's all.

    ReplyDelete

COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.